SVM Classifier WBCD data¶
In [1]:
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
%matplotlib inline
In [9]:
df = pd.read_csv("wbcd.csv")
In [7]:
df.head()
Out[7]:
In [12]:
df.corr()
Out[12]:
In [13]:
df.info()
In [15]:
df.shape
Out[15]:
In [16]:
df.describe()
Out[16]:
In [17]:
round(df.describe(),2) # round up to 2 zeros
Out[17]:
In [18]:
sns.pairplot(df,hue='diagnosis',palette='Set1')
Out[18]:
In [19]:
df['diagnosis'].value_counts() # value count of target variable size
Out[19]:
In [20]:
sns.countplot(data=df, x='diagnosis')
Out[20]:
lets Normalize our data Featurs Scalling of our data¶
In [21]:
from sklearn.preprocessing import StandardScaler
In [22]:
scaler = StandardScaler()
In [23]:
scaler.fit(df.drop('diagnosis',axis=1))
Out[23]:
In [24]:
scaled_features = scaler.transform(df.drop('diagnosis',axis=1))
In [29]:
df_feat = pd.DataFrame(scaled_features,columns=df.columns[:-1])
df_feat.head()
Out[29]:
Train test split¶
In [30]:
from sklearn.model_selection import train_test_split
from sklearn import metrics
In [31]:
X = scaled_features
y = df['diagnosis']
In [32]:
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)
SVM Classifiers¶
In [33]:
#import SVC classifier
from sklearn.svm import SVC
#import metrics to compute accuracy
from sklearn.metrics import accuracy_score
In [34]:
#initiacte classifier
svc = SVC()
In [35]:
svc.fit(X_train,y_train)
Out[35]:
Prediction and Evaluation¶
In [36]:
predictions = svc.predict(X_test)
In [37]:
from sklearn.metrics import classification_report,confusion_matrix
In [38]:
print(confusion_matrix(y_test,predictions))
In [39]:
print(classification_report(y_test,predictions))
In [40]:
print('Model Accuracy Score: {0:0.4f}'.format(accuracy_score(y_test,predictions))) # Accuracy predictions
Create Classification SVM classification linear object¶
SVM classifiction linear Kernel object
In [42]:
model_linear = SVC(kernel = "linear")
In [43]:
model_linear.fit(X_train,y_train)
Out[43]:
Prediction and Evaluation¶
In [44]:
pred_test_linear = model_linear.predict(X_test)
In [45]:
np.mean(pred_test_linear==y_test)
Out[45]:
Kernel poly¶
In [46]:
model_poly = SVC(kernel = "poly")
In [47]:
model_poly.fit(X_train,y_train)
Out[47]:
Prediction and evaluation¶
In [48]:
pred_test_poly = model_poly.predict(X_test)
In [49]:
np.mean(pred_test_poly==y_test)
Out[49]:
Kernel RBF¶
In [50]:
model_rbf = SVC(kernel = "rbf")
In [51]:
model_rbf.fit(X_train,y_train)
Out[51]:
Predicition and Evaluation¶
In [52]:
pred_test_rbf = model_rbf.predict(X_test)
In [53]:
np.mean(pred_test_rbf==y_test)
Out[53]:
Kernel Sigmoid¶
In [54]:
model_sig = SVC(kernel="sigmoid")
In [55]:
model_sig.fit(X_train,y_train)
Out[55]:
Predicition and Evaluation¶
In [56]:
pred_test_sig = model_sig.predict(X_test)
In [57]:
np.mean(pred_test_sig==y_test)
Out[57]:
In [ ]: